Stabilizing dynamics recur across the AMOC–cryosphere system—the circulation, ice, sea-ice and permafrost components examined in the commentary—and may alter the pace, extent or occurrence of modeled climate change, according to a new commentary. But the author does not say these effects outweigh destabilizing influences. The message is narrower and important: tipping-risk assessments should look for both the processes that accelerate change and those that slow or delay it.
The manuscript is an arXiv preprint marked v1 and dated 26 Aug 2026. It reviews cited climate-model simulations, feedback decompositions and tipping-interaction assessments involving the AMOC, Greenland Ice Sheet (GIS), West Antarctic Ice Sheet (WAIS), Arctic sea ice, permafrost and Antarctic sea-level response. No new data or original experimental sample were generated; the model output shown in its figure comes from Sinet et al. (2025), whose corresponding data are publicly available.
Examples of a modeled brake
Some of the clearest examples concern a weakening AMOC. In cited simulations, AMOC weakening was associated with 36% and 22% lower decline rates for Arctic sea-ice area and volume, respectively, over 1980–2020. Put simply, the modeled losses were slower during that interval. Across other cited experiments, the review reports 26–57% lower permafrost carbon loss during a temporary AMOC slowdown.
Greenland offers another clear example of a modeled delay. In cited coupled simulations, Greenland Ice Sheet melting was described as effectively halted at about 3°C of global warming after an AMOC collapse, while disintegration was substantially delayed at higher warming levels.
Antarctica shows why context matters
Antarctica adds a less uniform picture. A cited model study found little total Antarctic ice-volume response for several centuries after AMOC shutdown, no modeled WAIS destabilization, and longer-term subsurface cooling associated with lower basal melt.
When Antarctic freshwater scenarios were examined, the AMOC response did not point in one direction. The review describes cases ranging from slight strengthening to weakening, delayed weakening and re-establishment from an off state. Resilience varied with the timing and duration of ice-sheet freshwater input, including cases involving WAIS-tipping meltwater and cases with no AMOC collapse.
That dependence on circumstances is central to the commentary. Within the AMOC, contributions linked to heat movement and freshwater transport by ocean gyres were described as stabilizing in some states, while the feedback linked to salt transport could be stabilizing or destabilizing depending on background freshwater transport. The sign of an interaction, in other words, is not always a fixed property of the two components alone.
Stabilization inside the ice
Stabilization can also arise inside the ice sheets themselves. In cited simulations, Antarctic grounding-line retreat—the retreat of the boundary between grounded and floating ice—occurred 50–130 years later, and the Antarctic sea-level contribution was 9–23% lower by 2500. The review also describes a weakening of Greenland’s melt–elevation feedback.
A case for changing the risk picture
In one cited assessment, destabilizing interactions were more common: 15 of 20 tipping-element interactions were classified as destabilizing, four as stabilizing and one as unclear or competing. The author proposes three priorities for fuller risk assessment: systematic attention to stabilizing dynamics, a context-aware way to frame interactions, and methods to detect, anticipate and attribute stabilization.
Those methods include causal-inference approaches and physics-based feedback decompositions, which the review highlights as promising tools for finding stabilizing effects and tracing how they operate. The conceptual goal is to record the conditions behind an interaction rather than assign it one permanent sign when its response can change with state, timing or duration.
The conclusion is deliberately limited. Stabilizing dynamics are recurrent and consequential across and within the AMOC–cryosphere system, but the commentary does not determine whether they outweigh destabilizing influences. It offers a case for changing how tipping risk is assessed, not a final verdict on the system’s net direction.
Paper data and sources
Original title: A place for stabilization alongside tipping cascades: the AMOC-cryosphere system
Authors: Sacha Sinet
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-26
DOI: Not available
Original paper · Full text